基于期望横摆角速度的视觉导航智能车辆横向控制

Vision Guided Intelligent Vehicle Lateral Control Based on Desired Yaw Rate

针对采用传统位置偏差控制方法的车道保持系统横向控制精度不高以及鲁棒性差等问题,提出一种跟踪期望横摆角速度的车辆横向控制方法。在车辆当前行驶位置和道路预瞄点之间实时规划逼近目标路径的虚拟路径,同时分析当前时刻车辆以无偏差形式沿此虚拟路径行驶时决定车辆行驶位置的横摆角速度及速度之间的关系。结合车辆道路相对位置及车身状态信息,设计期望横摆角速度生成器。基于7自由度非线性车辆动力学模型,设计滑模控制器跟踪期望横摆角速度,使得车辆稳定地跟踪目标路径。根据车道线宽度和边缘点数量统计进行边缘检测,能有效识别模糊车道边缘和抑制噪声,并通过对消失点的检测来有效去除非车道线的干扰。仿真及试验结果表明,与传统的位置偏差控制方法相比,期望横摆角速度法不仅能提高车辆横向控制的精确性且跟随偏差随车辆速度及道路曲率的变化波动范围小,具有很好的鲁棒性和自适应性。

According to the low lateral control accuracy and low robustness of lane keeping system based on typically position error control,a lateral control algorithm based on tracking desired yaw rate is proposed.A virtual path is planned between vehicle and preview target point,yaw rate and velocity are analyzed while a vehicle tracking the virtual path with zero tracking error,a desired yaw rate generator is designed based on the body status and correspond positions of vehicle and preview target point.A yaw rate tracking controller is designed to track the desired yaw rate based on vehicle seven-degree-of-freedom nonlinear dynamic model,and vehicle can run according to the virtual path accurately in the end.Lane edge detection algorithm based on lane width and statistical edge points can detect vague lane and eliminate noise,vanishing point is detected to filter the straight line which are not lane line.Simulation and experiment studies shows that the proposed lateral control algorithm not only can effectively control vehicle to run according to the pre-given path perfectly under different velocities and road curvature but also have good robustness and adaptability.